Collaborations to provide bio-statistical expertise continued on five major epidemiologic studies within DCEG: (1) the Agricultural Health Study, a prospective study of a cohort of 90,000 adults living on farms in Iowa and North Carolina to investigate potential disease risks associated with agricultural exposures; (2) the ALTS trial, a randomized trial to examine whether human papilloma virus DNA type can be used to determine the appropriate treatment for women with abnormal PAP smears; (3) a case-control study of brain cancer to examine a variety of occupational exposures as well as cellular phone use; (4) a case-control study of testicular germ cell tumors; and (5) a cohort study of Radiologic Technologists. Collaboration continued on a prospective study of possible cancer risk in men exposed to radar during military service in the 1950s, on analyses of lycopene and selenium intake in a case-control study of prostate cancer, and on a case-control study of non-Hodgkins lymphoma to investigate possible causes for the unexplained secular increase in rates. Studies of calendar period and birth cohort trends in kidney cancer mortality and incidence rates and prostate cancer mortality rates in the United States are ongoing. A follow-up study of retinoblastoma survivors continues, with current emphasis on a significant excess of uterine tumors in female germline mutation carriers. Additional projects and studies for which assistance continues to be provided include: an Italian case-control study of melanoma evaluating the possible role that differences in DNA-repair capacity might play in determining risk; a study examining the clinical significance of different mutations in the RB1 gene in hereditary retinoblastoma patients; a study to assess the ability to identify determinants of pesticide exposure from videotapes of pesticide application procedures. Collaboration continued in evaluation of the effectiveness of the American Stop Smoking Intervention Study (ASSIST) in reducing smoking levels in states given tobacco control resources. A two-stage analysis of smoking prevalence was developed to evaluate the effect of the ASSIST funding. A repeated measures random effects model is being used to determine if tobacco sales declined faster in ASSIST states than in the non-ASSIST states. Analyses of the NHANES I Epidemiologic Follow-up Study cohort are being conducted to examine associations between physical activity and breast cancer risk, and between aspirin intake and cancer and cardiovascular mortality. The mortality follow-up to the Second National Health and Examination Study is being used to study associations between blood lead levels and aspirin intake on cancer and cardiovascular mortality. Analyses of food frequency data in the 1987 and 1992 Epidemiologic Supplements to the National Health Interview Survey (NHIS) and the mortality follow-up to the NHIS are examining associations between diet quality and carotenoid intake and cancer mortality.